Only few systematic studies are available on the status of sleep deprivation therapy in the overall treatment regimen of depressive patients. 51 patients suffering from a major depressive episode (ICD-10) were randomly allocated to 4 weeks' treatment with amitriptyline (150 mg/day) or to a combination of amitriptyline with six partial-sleep-deprivation treatments late in the night (at 4-5 day intervals). According to observer rating (Hamilton Rating Scale for Depression, 21- and 10-item version), a highly significant amelioration was recorded in both groups until the 14th day of treatment. A further improvement occurred, however, only in those patients treated with both antidepressants and sleep deprivation. Hence the response rate ( > or = 50% HAMD reduction) after 4 weeks' treatment was distinctly more favourable in this group than in those patients under pharmacotherapy alone. The superiority of the combined therapy cannot be confirmed statistically by self-rating (Befindlichkeitsskala: von Zerssen; Visual Analogue Mood Scale). The immediate antidepressive effect of sleep deprivation diminished in the course of the sleep deprivation series. The response to the first sleep deprivation was a predictor neither for the response to further sleep deprivation treatments nor for the overall treatment outcome.